Predictive Model for the Diagnosis of Benign/Malignant Complex Cystic and Solid Breast Nodules

Abstract

Purpose: To develop an ultrasound predictive model to differentiate between benign and malignant complex cystic and solid nodules (C-SNs). Methods: A total of 211 patients with complex C-SNs rated as American College of Radiology Breast Imaging Reporting and Data System (ACR BI-RADS) category 4 or 5 on the ultrasound reports were included in the study, from June 2018-2021. Multivariate stepwise logistic regression analysis was used to establish a predictive model, based on clinical and ultrasound features. The diagnostic performance of the model was evaluated by the area under the curve (AUC) of the receiver operating characteristic curve. Results: A total of 109 breast nodules, including 74 benign nodules (67.89%) and 35 malignant nodules (32.11%), were detected by surgical pathology or puncture biopsy. Multivariate analysis showed that the blood flow (BF) of complex C-SNs (p = 0.03), cystic fluid transmission (p = 0.02), longitudinal diameter (p < 0.001), and age (p = 0.03) were independent risk factors for malignant complex cystic and solid breast nodules. The ultrasound model equation was Z = −12.14 + 2.24 × X12 + 1.97 × X20 + 0.40 × X7 + 0.11 × X0; M = e z 1 + e z (M is the malignancy score, e = 2.72). The area under the curve (AUC) was 0.89, which indicated good predictive utility for the model. Conclusions: A prediction model incorporating major risk factors can predict the malignant C-SNs with accuracy.